Ethnopolitics, Vol. 4, No.

4, 359– 363, November 2005

Introduction: Elections in Divided

Societies

FLORIAN BIEBER & STEFAN WOLFF


European Centre for Minority Issues, Flensburg, Germany;  Department of European Studies, University of
Bath, UK

Divided societies around the world require specific institutional arrangements, including
electoral rules, to accommodate diversity of interest in ways that can deal with social
divisions in a non-violent manner. What might sound like a truism is still frequently
challenged in discussions on institutional design for divided societies. While scepticism in
academia and among policy makers often focuses on whether democracy is at all possible
in divided societies, a large number of countries have been forced to meet this particular
challenge in recent decades. In some cases, such as Belgium, long-established democracies
have experienced increasing division along linguistic (or ethnic) lines. In others, such as
Macedonia, the transition to democracy has occurred under conditions of deep ethnic division
and a long-standing antagonism between different communities with very distinct visions of
the future of the state in which they live. A third category of cases is constituted by countries
in which hegemonic parties in the past have dominated and at times constrained democratic
competition, such as in Singapore or in Rwanda, but have subsequently moved towards more
competitive elections. The marked divisions along ethnic, religious or racial lines not only
inform today’s voters’ choices at the ballot box but they also frequently precede the specific
electoral rule and political system in place today.
The conceptual discussions and case studies that follow this introduction focus on the
evolution of electoral systems in seven countries around the world, from Belgium to
Trinidad and Tobago. In some cases, such as Rwanda, experiments with multiparty
elections in the post-genocide context of the country are recent, whereas in other cases,
such as Guyana, elections discussed here date back several decades. The cases
demonstrate that the ethnopolitical dynamics of elections are best understood in the
context of the evolution of electoral systems, and not just by a snapshot analysis of
their technical design. Understandably, much discussion on electoral systems in recent
years has focused on the institutional design component: academic and policy-oriented
research has made significant contributions to devising new electoral systems to
‘manage’ democracy and diversity in post-conflict divided societies, ranging from

Correspondence Address: Florian Bieber, Svetog Save 34/10, 11000 Belgrade, Serbia and Montenegro. Email:
bieber@gmx.net; Stefan Wolff, Department of European Studies, University of Bath, Bath BA2 7AY, UK.
Email: S.Wolff@bath.ac.uk

1744-9057 Print=1744-9065 Online=05=040359–05 # 2005 The Editor of Ethnopolitics

DOI: 10.1080=17449050500382872
360 F. Bieber & S. Wolff

South Africa to Bosnia and Iraq. Whereas the choice of the electoral system is obviously
crucial in setting the framework for the emergence of a stable democratic process and its
ability to manage existing divisions peacefully within such societies, only the passing of
time can provide more insight into the effectiveness of electoral rules and the evolving
relationship between democracy and identity. In fact, the ‘design’ of electoral systems,
like that of other institutions, is driven not only be intention but often also by accident
or evolution and the consequences of social and political change (Goodin, 1998,
pp. 24– 30).
The way in which group interests are represented in any political system is also the
result of a number of other factors often not related to ethnicity. Adjustments to the elec-
toral system frequently have a profound relevance for minorities, although these changes
are made on the basis of other considerations. In the case of Singapore the expansion of
Group Representation Constituencies in the 1990s, as Eugene Tan details, appears at
first sight to be a measure directed at enhancing the representation of smaller communities,
in particular Malays. In fact, these constituencies primarily served to consolidate the
political control by the dominant People’s Action Party. Similarly in Macedonia,
the majoritarian electoral system in use until 2002 was primarily motivated by the
incumbents’ desire to secure political dominance.
The two predominant schools in conflict resolution today—integrative and consocia-
tional power sharing—take very distinct views on which electoral systems stand the
best chance of contributing to the successful management of conflict in deeply divided
societies. As the debates on institutional and electoral system design are interlocked,
there often appears to be a choice only between a proportional system, advocated by
Arend Lijphart and other proponents of consociational systems, and ‘integrative’ electoral
systems, such as alternative votes, suggested by Donald Horowitz and others (Sisk, 1996,
pp. 27 –45). As the following contributions demonstrate, this line of debate remains
important, but constitutes at best only the starting point for a number of other key
issues involved in addressing ethnic diversity in electoral systems.1 In his contribution
on Slovakia and Macedonia, Eben Friedman argues that the choice of proportional
representation in majoritarian electoral systems for divided societies is not a foregone
conclusion. While more effective in securing the representation of larger groups, such
as Magyars in Slovakia or Albanians in Macedonia, proportional representation might
disadvantage communities which are either small or which lack the necessary degree of
political mobilization, such as the Roma in these two countries and elsewhere in Central
and Southeastern Europe. Small minorities often find it extremely difficult to achieve
representation in parliament through standard electoral rules. If the goal is indeed to
assure the representation of members of smaller groups in parliament, this might require
special minority-friendly rules, such as reserved seats or separate electoral rolls.
A related key challenge in electoral systems which seek to ensure the representation of
groups and minorities is the nature according to which ethnicity, race or religion is
represented. In the case of Belgium, as Jean-Benoit Pilet emphasizes, the determination
of votes along linguistic lines is unacceptable, as this is viewed to foster a potentially divi-
sive ‘linguistic nationality’. In Singapore, on the other end of the spectrum, citizens are
assigned their racial identity at birth. Predetermining the relations between individuals
and their ethnic, racial and national identities naturally constrains the range of electoral
systems available for representing groups. Even more so, such policies also limit the
degree of subjective choice any given electoral system might offer. As the Bosnian
 Introduction: Elections in Divided Societies 361

Constitutional Court pointed out in a landmark ruling in 2000 in regard to the predetermi-
nation of ethnicity, “the Belgian system does not exclude per se the right to stand as a
candidate solely on the ground of language. Every citizen can stand as a candidate,
but has—upon his election—to decide whether he will take the oath in French or in
Flemish. It is therefore the subjective choice of the individual candidate whether to take
the oath in French or in Flemish and thereby to ‘represent’ a specific language group”
(Constitutional Court, 2000, para. 120, original emphasis). The dilemma between allowing
candidates and voters to determine their ethnic, racial or religious identity on the basis of
subjective criteria or defining a more rigid and possible even quasi ‘objective’ system for
determining affiliation to a particular group is a reccurring problem in electoral systems
which give representation to particular identities. The fundamentally subjective nature
of such identities suggests that any predetermined identification without the consent of
the individual, be it a candidate or a voter, is inappropriate in a liberal democracy. At
the same time, a system allowing full, unchecked subjective choice might be equally
problematic. Consider, for example, the well documented case of Hungary’s local
minority councils. All citizens have been able to vote for these minority councils during
municipal elections, resulting in a large number of impostors reaping the benefit of
being elected to represent a minority they clearly have no link to. Furthermore, voting
across the community divide does not always help moderation, as the case of Brussels
illustrates. In the only region of Belgium which allows voting across linguistic lines,
the extremist Vlaams Blok has been more successful in wooing Francophone voters on
its anti-immigrant platform than more moderate parties.
The nature of representation of ethnic and other indentities also informs the larger func-
tion of electoral systems or rules which address diversity. Matthijs Bogaards has distin-
guished between electoral systems which block ethnicity, those which aggregate and
those which translate ethnicity (2004, pp. 250 – 256). Translating ethnicity is clearly
linked to proportional representation, whereas aggregation transforms ethnic identity
through electoral rules such as the alternative vote. Blockage finally suppresses the
expression of ethnicity in the electoral system. One-party systems, such as in Rwanda
and Singapore, in particular fall into this category. Interesting, however, is the divergence
between the two. As David Emmanuel Kiwuwa demonstrates, Rwanda qualifies as a
classical electoral system aimed at blocking ethnic representation with the goal of down-
playing the division between Hutus and Tutsis as a factor in the country’s post-genocide
politics and possibly disguising the ethnic bias of the governing party. In Singapore, on the
other hand, Eugene Tan argues that ethnic polarization has been blocked or at least heavily
discouraged, but this dominant one-party system has also sought to institute a system of
translating the diversity in the population at large into representation in parliament. In
fact, these two divergent strategies can be found widely in many countries in Africa
and Asia. The suppression of ethnicity as a politically relevant cleavage bears a number
of obvious problems, as detailed by Kiwuwa for the case of Rwanda. First, it easily
allows a hegemonic party to stave off competition by accusing it (rightly or wrongly) of
being organized along ethnic lines. This danger particularly poses itself in majority–
minority situations were majority ethnopolitical parties are accepted as regular political
parties, whereas similar parties among minorities are defined as a potential threat and a
target for banning. Second, such a ban is unlikely to render ethnicity irrelevant. In fact,
such measures frequently accentuate the importance of ethnicity outside official
politics, especially when ethnic cleavages had been previously instrumentalized.
362 F. Bieber & S. Wolff

As a result, ethnicity might matter, even if officially discouraged, and challenge the
legitimacy of such policies and the regimes who propose them. The approach in
Singapore, on the other hand, reflects the equally common policy not to ban ethnic
representation as such, but to encourage cross-ethnic coalitions in a restrictive framework.
As in the case of Rwanda, this approach has been closely associated with one hegemonic
political party. The predominant Group Representation Constituencies require multiracial
candidate lists, which discourage mono-racial parties. Similarly, presidential elections in
Nigeria and Kenya require candidates to gather a minimum number of votes (and/or
signatures for registration) from different regions to ensure the geographic (read ethnic)
representativeness of the candidate. The case of Lebanon, which also has multi-member
multi-religious electoral units, suggests that such systems do not per se benefit cross-
ethnic parties or coalitions because of tokenism and temporary cross-community alliances.
An alternative approach in the case of dominant parties are consociational features within
the dominant party, as in the case of Singapore, but arguably also to be found in the ANC
in South Africa (Bogaards, 2005, pp. 164– 183).
In addition to the function of the electoral rules, the factor time is crucial in disentan-
gling electoral rules and electoral systems. As Taagepera has argued, electoral systems
are only the process of long-term evolution and ‘appropriation’ of a set of electoral
rules into the political system (2002, pp. 248– 264). As such, the electoral rules might
acquire different meanings over time with the development of the electoral system. In
Belgium the electoral rules changed much later than the system itself. As a result, the
same electoral rules governed two different political systems: one in which only a few
seats were won by separate linguistic parties and most electoral contestation took place
between the ‘spiritual families’; and, later, a system where no party cut across the linguis-
tic boundary and the division of the country shifted from one based on ideology and
religion to one based on language. The cases of Guyana and Trinidad and Tobago illustrate
the same type of development, if only in the reverse direction. Here, as Sara Abraham
demonstrates, multiracial parties and coalitions won in both countries in a setting where
otherwise mono-racial parties had prevailed (before and after). Unlike in the case of
Belgium, where the transformation towards a more ethnicized party system has been
lasting, the switch to multiracial parties has been temporary. As Abrahams shows, this
temporary reorientation of the electorate has been closely linked with larger economic
transformations and major crises in the political systems (decolonization/decline of
dominant party) which called for a broader coalition. The trajectory of the electoral
systems in the countries under discussion here thus suggests that emergence of a system
focused on ethnic lines can be the result of a gradual transformation, while the reduced
rule of ethnicity is either of a temporary nature (Trinidad and Tobago, Guyana) or the
result of explicit blockage (Rwanda).
A key lesson from these cases of elections in diverse societies is related to the develop-
ment of electoral systems. Whether in established democracies like Belgium or in more
recent democracies such as Macedonia or in countries like Rwanda which are just begin-
ning to see multiparty elections (albeit severely constrained), electoral systems have been
in flux and adjusted to the demands and needs of the societies’ mobilized ethnopolitical
groups. The changing nature of electoral systems in diverse societies suggest that there
is neither one clear approach for all divided societies, nor is there one set of electoral
rules which will not require adjustment over time to reflect the changing relationship
between identity, society and democracy.
 Introduction: Elections in Divided Societies 363

Note
1. For a more detailed analysis of these two interlocked debates on institutional and electoral system design, see
Wolff (2005).

References
Bogaards, M. (2004) Electoral system and the management of ethnic conflict in the Balkans, in: A. Mungiu-
Pippidi & I. Krastev (Eds), Nationalism after Communism: Lessons Learned, pp. 249–268 (Budapest:
Central European University Press).
Bogaards, M. (2005), Power sharing in South Africa: the African National Congress as a consociational party?, in:
S. Noel (Ed.), From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies,
pp. 164 –183 (Montreal: McGill-Queen’s University Press).
Constitutional Court of Bosnia and Herzegovina (2000) Partial Decision, Case No. U 5/98-III, Request for evalu-
ation of constitutionality of certain provisions of the Constitution of Republika Srpska and the Constitution
of the Federation of Bosnia and Herzegovina.
Goodin, R. E. (1998) Institutions and their design, in: Goodin (Ed.), The Theory of Institutional Design,
pp. 24 –30 (Cambridge: Cambridge University Press).
Sisk, T. (1996) Power Sharing and International Mediation in Ethnic Conflict (Washington, DC: US Institute of
Peace).
Taagepera, R. (2002) Designing electoral rules and waiting for an electoral system to evolve, in: A. Reynolds
(Ed.), The Architecture of Democracy. Constitutional Design, Conflict Management, and Democracy,
pp. 248 –264 (Oxford: Oxford University Press).
Wolff, S. (2005) Electoral systems design and power-sharing regimes, in: I. O’Flynn & D. Russell (Eds), Power-
sharing: New Challenges for Divided Societies, pp. 59 –74 (Ann Arbor, MI: University of Michigan Press).